Placebos: when?

[Oualawouzou]

Hello everybody,

I'll try to make this short and to the point.

I was involved recently in a brief online debate regarding a new HIV vaccine being tested on prostitutes in Dominican Republic. The discussion was semi-hijacked to discuss the use of placebos.

My position was that placebos were absolutely necessary in clinical trial for a variety of reasons:
-remove any and all bias in the data collection and analysis;
-prevent discrepancies in the habits of subjects, doubly important in the present situation as contacting HIV is highly dependant (sp?) on the subject's actions. You don't want someone who knows he got the vaccine to act differently from someone who knows he hasn't got it, to properly gauge the efficacity of the vaccine;
-create a control group that you can monitor just as closely as the one who gets the vaccine; without it, you can't make an accurate evaluation of the efficacity of the vaccine (does it not work? Does it block all infections? Does it merely reduce the risks of infection?).

My opponent stated placebos were not necessary because:
-placebos are used only in tests where a qualitative evaluation takes place (example: pain) rather than a quantitative one;
-placebos are forbidden of use when the subject could be encouraged directly or indirectly to engage in risky behavior (stop following other treatment or, in the case at hand, engage in risky sex/other dangerous ways to contract HIV);
-there is no need to create a control group as they simply need to look at information already collected in the past years regarding HIV among sex workers in the D.R.

So... I did some research and found plenty of interesting information about placebos, but nothing to tell who was right (totally or in part). Can any of you help?

Thanks!

 

[drkitten]

My opponent stated placebos were not necessary because:
-placebos are used only in tests where a qualitative evaluation takes place (example: pain) rather than a quantitative one;
-placebos are forbidden of use when the subject could be encouraged directly or indirectly to engage in risky behavior (stop following other treatment or, in the case at hand, engage in risky sex/other dangerous ways to contract HIV);
-there is no need to create a control group as they simply need to look at information already collected in the past years regarding HIV among sex workers in the D.R.

Your opponent is wrong. In fact, his first two points contradict each other. The point of placebo control is to control for all the psychological factors involved in treatment, such as the possibility that people will change their behavior simply as a result of receiving medical treatment. One possible outcome of the vaccine, for example, is that prostitutes will be acting in a more risky fashion because they believe that they are protected -- but another possibility is that they will actually be acting in a more risk-averse fashion because they've learned (or been reminded) of the possible negative consequences of risky behavior. It's exactly these types of confounds that are the reason placebo controls exist.

Now, he's certainly right that it would be inethical to withhold effective treatment. In cases where there is an effective treatment available, the usual procedure would be to give the experimental treatment to one group and the normal (effective) treatment to the control group instead of giving sugar pills or something. But that doesn't apply in this case, since as far as I know there is no "effective" vaccine to withhold.

 

[Babbylonian]

It seems like the purpose of a placebo in the described case would be to measure the potential side effects of the vaccine (by patient report and physical examination) at the same time as measuring the efficacy of the vaccine (by testing the blood for HIV antigens/antibodies). Given proper education of the participants, particularly on the number of other potentially deadly STDs that can be contracted through unsafe behavior, the behavioral change should be minimal, thus obviating the ethical problem of "encouraging" risky behavior. So, yes, unless the question is narrowed to only the efficacy of the vaccine, a placebo control would seem appropriate...to this layperson, anyway.

 

[Dark Jaguar]

It's not like these people wouldn't be informed they have only a 50% chance of actually getting the experimental treatment beforehand, right?

 

[Art Vandelay]

You seem to be confusing control group and placebo. The control group usually has a placebo, but that doesn't mean that they are they same concept.

One possible outcome of the vaccine, for example, is that prostitutes will be acting in a more risky fashion because they believe that they are protected -- but another possibility is that they will actually be acting in a more risk-averse fashion because they've learned (or been reminded) of the possible negative consequences of risky behavior. It's exactly these types of confounds that are the reason placebo controls exist.

On top of that, there's a selection bias, in which people who volunteer for a study are probably statitically different from the general population, so comparing the results in the former to the latter is not legitimate.

But that doesn't apply in this case, since as far as I know there is no "effective" vaccine to withhold.

That, of course, raises the issue of what if there were: is it unethical to withold a proven treatment to test an unknown treatment?

 

[drkitten]

You seem to be confusing control group and placebo.

Art, you don't need to prove to me that you have reading skills that would embarrass an eight-year old. Your posts on the politics forum have proven that amply to me.

The control group usually has a placebo, but that doesn't mean that they are they same concept.

They're not, which is why I specifically mentioned that it's usually inethical to give placebos to a control group if there is an effective treatment already available..

That, of course, raises the issue of what if there were: is it unethical to withold a proven treatment to test an unknown treatment?

Hardly an issue. No, it's not as I specifically addressed in the post you couldn't read because of all the big words.

Please remember your membership agreement regarding civility toward other members... this kind of interaction distracts from the topic and lowers the quality of posting in a thread.

Replying to this modbox in thread will be off topic  Posted By: jmercer

 

[Ben Tilly]

My position was that placebos were absolutely necessary in clinical trial for a variety of reasons:
-remove any and all bias in the data collection and analysis;
-prevent discrepancies in the habits of subjects, doubly important in the present situation as contacting HIV is highly dependant (sp?) on the subject's actions. You don't want someone who knows he got the vaccine to act differently from someone who knows he hasn't got it, to properly gauge the efficacity of the vaccine;
-create a control group that you can monitor just as closely as the one who gets the vaccine; without it, you can't make an accurate evaluation of the efficacity of the vaccine (does it not work? Does it block all infections? Does it merely reduce the risks of infection?).

You're right. Let me add the fact (not relevant for HIV, but relevant in many other cases) that frequently someone who believes they will show improvement, actually does. In fact it has been demonstrated that even if the patients aren't told whether they are getting a placebo, but the experimenter knows, that the patient outcomes will differ. Which is why, when possible, you want experiments to be conducted "double blind". Meaning that neither the patient nor anyone dealing with that patient knows who is in the control group.

My opponent stated placebos were not necessary because:
-placebos are used only in tests where a qualitative evaluation takes place (example: pain) rather than a quantitative one;

Absolutely wrong. Placebos are frequently used for tests of all sorts of quantitative things. Such as how effective drugs are, where effective may be something concrete like "didn't have repeat heart attack within period X".

-placebos are forbidden of use when the subject could be encouraged directly or indirectly to engage in risky behavior (stop following other treatment or, in the case at hand, engage in risky sex/other dangerous ways to contract HIV);

Who forbids this? As long as the patients are aware that they might be getting a drug or might be on a placebo, I can't see an ethics committee objecting.

-there is no need to create a control group as they simply need to look at information already collected in the past years regarding HIV among sex workers in the D.R.

That's not a valid control group. You never know how different people are who are self-selecting to be part of your group. Furthermore behaviours and risks in a population change over time. Was a reduction in cases due to your treatment or was it due to some kind of outreach program changing everyone's risks?

Hopefully that helps.

Cheers,
Ben

 

[Oualawouzou]

Thanks for your replies all. It's good to know I was about right, but I'm even more glad to know how to counteract the opposing arguments next time.

 

[Art Vandelay]

Art, you don't need to prove to me that you have reading skills that would embarrass an eight-year old. Your posts on the politics forum have proven that amply to me.

And you don't need to prove to me that you're an ass. If you had actually paid attention to my posts, rather than making a habit of merely posting knee-jerk responses that show that you aren't even bothering to listen to what I'm saying, you might have noticed that I tend to make responses to other posters after quoting them, and anything which is not preceded by a quote is usually in response to the OP. But thanks for showing, to anyone who was previously unaware, that you're the sort of person who leaps to conclusions about other people, and then makes vicious, unprovoked attacks that are based on those conclusions.

Hardly an issue. No, it's not as I specifically addressed in the post you couldn't read because of all the big words.

You merely mentioned what the usual procedure is. You didn't address the ethical implications.

BTW, I checked a few dictionaries, and "inethical" wasn't in any of them.